What is DaaS (data as a service)?

The name borrows from cloud computing, where NIST's widely cited definition sets out service models such as software as a service. The common thread is that the customer uses something the provider runs, rather than installing or keeping it. With DaaS, the thing is data: company records, prices, weather, locations, risk scores. The provider collects, cleans and updates it; the customer queries what it needs when it needs it.

Beware a naming clash. In IT, DaaS also stands for desktop as a service, virtual desktops hosted in the cloud. Search results mix the two. This page is only about data.

• API: the customer's software requests records as needed, often billed by call or by tier.
• Scheduled feed: new or changed records land in the customer's storage on a schedule.
• Cloud data share: the customer queries the provider's live tables inside the same warehouse, as Snowflake's sharing and AWS Data Exchange's Redshift and S3 options allow.
• Web application: a search interface on top of the data, which blurs into software as a service.

Suppose an agricultural cooperative runs a network of private weather stations across its members' farms. A crop insurer first asks to buy five years of history. That is a one-off dataset licence. Then the insurer asks for daily updates to price new policies. Now the cooperative is considering DaaS: it needs reliable uptime, an API or a daily file drop, a support contact when a station fails, and a subscription agreement instead of a single payment.

In this illustration, the cooperative may not want to run a service. A broker could help it compare options: sell the history once and stop, partner with a provider that already runs weather feeds, or list the feed on a cloud marketplace that handles delivery. Each route keeps the cooperative as the source while shifting the operational burden differently.

• Rights: a subscription is still a licence. Check permitted uses, caching limits and what must be deleted on cancellation.
• Freshness and coverage: how often records change and how gaps are reported.
• Service levels: uptime commitments and what happens when the API fails.
• Lock-in: schemas and identifiers differ between providers, so switching can be costly.
• Personal data: many DaaS products in sales and marketing contain contact details, which bring privacy and registry duties.

Most people searching for DaaS want a provider to buy from. We don't sell data feeds. This site is for people who broker datasets between companies, including deciding whether a seller's data should become a service, a marketplace listing or a single licence.